The present invention relates to a network system, method and computer program constructed in a Local Area Network (LAN). More particularly, the present invention relates to an automatic address distributing method, computer program and system having a function for distributing addresses each serving as identification information of each terminal connected to the LAN.
With the development of network systems, there have been encountered problems that the amount of work load in system construction and management increases and addresses each serving information for uniquely identifying each terminal connected to a network become obsolete. In order to solve the problems, there have been proposed methods of automatically distributing an address which can be used only for a fixed period of time to each terminal connected to a network.
A representative one of the methods is a Dynamic Host Configuration Protocol (DHCP) described in the DHCP Network Working Group Requests for Comments (RFC) by R. Doms, IETF, Oct. 1993, information on standardizations open to the public. The DHCP allows the amount of work load in the system construction and management to be reduced and makes efficient use of addresses possible.
A point in the dynamic distribution of addresses according to the conventional technology to which attention should be paid is that the same address as one used for a terminal connected to the network must not be distributed to another terminal, that is, duplicate assignment of an address must be avoided. There are two techniques for addressing this problem.
According to a first technique addresses used in a network are examined in advance. The addresses found used in an existing network by the examination conducted in advance are never cataloged in a database used for storing addresses to be distributed by the server having a function for distributing addresses by the DHCP so as to avoid duplicate assignment of an address.
According to a second technique, in order to avoid duplicate assignment of an address by the DHCP, a packet is transmitted to an address to be assigned to a terminal prior to the distribution of the address. In this way, it is possible to determine if the address has been assigned to an existing terminal connected to the network.
The first technique has a problem that an additional work load for examining and cataloging addresses is introduced. Depending upon the size and complexity of the network, the problem caused by the additional work load may be severe.
The second technique has a problem that it may not be possible to examine whether or not a terminal having the same address as the one to be distributed exists due to the following reasons:
1) The terminal is in an off-line state due to, for example, the fact that the power supply thereof is turned off, PA0 2) Because of network congestion, the transmitted packet is lost, and PA0 3) The terminal is moved from the current network to another network with network information relevant to the terminal such as the address thereof left in the current network.
The shorter the period of time in which a distributed address can be used, in other words, the shorter the validity period of a distributed address, the higher the utilization efficiency of the address. However, a packet is required for requesting that the period of time or the validity period be extended. Transmission of such packets gives rise to a problem that the traffic of the network becomes excessively heavy.